Hugging Face's logo

Datasets:
Horama
/
wildlife-scrap

Tasks:
Image Classification
Languages:
English
French
Size:
100K<n<1M
Tags:
wildlife
animals
biology
zoology
inaturalist
flickr
License:
Dataset card Files
xet
 Community
1

You need to agree to share your contact information to access this dataset

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Log in or Sign Up to review the conditions and access this dataset content.

WOW – Wildlife Of the World

Multi-source wildlife image dataset covering 122 labels (species, sub-species and a few domestic categories) used at Horama for training the WOW classifier.

Each parquet shard is a HuggingFace Image feature so images decode straight to PIL.Image via load_dataset.

Quick stats

  • 134,281 images, 122 labels
  • last updated: 2026-04-29
  • sources: inaturalist, wikimedia, wikipedia, flickr, ddg

Splits (across all sources): train = 6,498 (70.0%) valid = 1,398 (15.1%) test = 1,386 (14.9%)

Source Images train valid test
inaturalist 117,478 0 0 0
wikimedia 7,521 0 0 0
wikipedia 0 0 0 0
flickr 0 0 0 0
ddg 9,282 6,498 1,398 1,386
TOTAL 134,281 6,498 1,398 1,386
Per-label image counts (122 labels, 134281 images) – click to expand
label specie inaturalist wikimedia wikipedia flickr ddg train valid test total
addax addax 746 54 0 0 0 0 0 0 800
agouti_azara agouti 2365 50 0 0 0 0 0 0 2415
alpaga alpaga 523 74 0 0 0 0 0 0 597
amazone_aourou amazone 840 38 0 0 0 0 0 0 878
ane_commun ane 2000 68 0 0 0 0 0 0 2068
ane_somalie ane 190 52 0 0 317 222 48 47 559
anoa anoa 80 46 0 0 249 174 38 37 375
antilope_cervicapre antilope 1485 70 0 0 0 0 0 0 1555
autruche autruche 1000 76 0 0 0 0 0 0 1076
bison_amerique bison 2018 67 0 0 0 0 0 0 2085
bison_europe bison 786 72 0 0 0 0 0 0 858
boeuf_ecosse boeuf 0 59 0 0 868 608 130 130 927
bongo bongo 257 53 0 0 344 241 52 51 654
cacatoes_rosalbin cacatoes 1465 75 0 0 0 0 0 0 1540
calao_abyssinie calao 1159 81 0 0 0 0 0 0 1240
calopsitte calopsitte 1127 81 0 0 0 0 0 0 1208
canard canard 1000 78 0 0 0 0 0 0 1078
capybara capybara 2000 79 0 0 0 0 0 0 2079
cariama_huppe cariama 1688 76 0 0 0 0 0 0 1764
casoar casoar 1843 53 0 0 0 0 0 0 1896
chat chat 1000 80 0 0 0 0 0 0 1080
cheval_comtois cheval 1000 63 0 0 0 0 0 0 1063
cheval_przewalski cheval 478 87 0 0 0 0 0 0 565
chevre_anglo_nubienne chevre 30 101 0 0 568 398 85 85 699
chevre_naine chevre 0 50 0 0 637 446 96 95 687
chien chien 1000 110 0 0 0 0 0 0 1110
chien_buisson chien_buisson 238 66 0 0 212 148 32 32 516
chouette_effraie chouette 2000 43 0 0 0 0 0 0 2043
coati_nez_blanc coati 1000 75 0 0 0 0 0 0 1075
cobe_croissant cobe 1000 56 0 0 0 0 0 0 1056
cobe_lechwe cobe 627 86 0 0 0 0 0 0 713
cochon_d_inde cochon 587 77 0 0 0 0 0 0 664
cochon_laineux cochon 1000 92 0 0 322 226 48 48 1414
colobe_guereza colobe 2334 63 0 0 0 0 0 0 2397
colombine_turvert colombine 1000 33 0 0 0 0 0 0 1033
corbeau corbeau 1000 73 0 0 0 0 0 0 1073
coyote coyote 1000 95 0 0 0 0 0 0 1095
daim_europe daim 2000 71 0 0 0 0 0 0 2071
dhole dhole 134 47 0 0 290 203 44 43 471
dromadaire dromadaire 871 77 0 0 0 0 0 0 948
eland_cap eland 2404 83 0 0 0 0 0 0 2487
elephant_afrique elephant 1000 56 0 0 0 0 0 0 1056
emeu emeu 1000 86 0 0 0 0 0 0 1086
fourmillier fourmillier 939 81 0 0 0 0 0 0 1020
gazelle_perse gazelle 1232 82 0 0 0 0 0 0 1314
gibbon_bonnet gibbon 329 45 0 0 254 178 38 38 628
girafe_kordofan girafe 189 1 0 0 189 132 29 28 379
gnou_bleu gnou 1000 69 0 0 0 0 0 0 1069
gnou_queue_blanche gnou 1183 71 0 0 0 0 0 0 1254
gorille_plaines gorille 712 71 0 0 0 0 0 0 783
grand_eclectus eclectus 384 24 0 0 299 209 45 45 707
grand_hocco hocco 754 28 0 0 0 0 0 0 782
grand_koudou koudou 2000 74 0 0 0 0 0 0 2074
grue_couronnee_grise grue 1114 30 0 0 0 0 0 0 1144
guepard guepard 1000 83 0 0 0 0 0 0 1083
hapalemur_lac hapalemur 918 2 0 0 208 146 31 31 1128
hippopotame hippopotame 2000 92 0 0 0 0 0 0 2092
hippotrague_noir hippotrague 1345 90 0 0 0 0 0 0 1435
hocco_daubenton hocco 33 1 0 0 379 265 57 57 413
hyene_tachetee hyene 1000 60 0 0 0 0 0 0 1060
kangourou_roux kangourou 1000 54 0 0 0 0 0 0 1054
lapin_geant_papillon lapin 2000 50 0 0 0 0 0 0 2050
lemur_couronne lemur 563 12 0 0 0 0 0 0 575
lemur_noir lemur 875 83 0 0 0 0 0 0 958
lemur_ventre_rouge lemur 640 22 0 0 0 0 0 0 662
lion_afrique lion 1000 58 0 0 0 0 0 0 1058
loriquet_arc_en_ciel loriquet 1000 85 0 0 0 0 0 0 1085
loup_arctique loup 296 47 0 0 396 277 60 59 739
loup_criniere loup_criniere 891 84 0 0 0 0 0 0 975
loup_mackenzie loup 348 30 0 0 285 199 43 43 663
loutre_asie loutre 1710 73 0 0 0 0 0 0 1783
lynx_carpates lynx 2057 64 0 0 0 0 0 0 2121
macaque_tonkean macaque 159 4 0 0 171 120 26 25 334
magabey_dore mangabey 859 86 0 0 207 145 31 31 1152
maki_catta maki 1358 80 0 0 0 0 0 0 1438
maki_vari_noir maki 1119 54 0 0 0 0 0 0 1173
maki_vari_roux maki 842 31 0 0 0 0 0 0 873
mara mara 1408 80 0 0 0 0 0 0 1488
milan_noir milan 1000 100 0 0 0 0 0 0 1100
moufette moufette 1000 66 0 0 0 0 0 0 1066
mouton_somalie mouton 0 7 0 0 487 341 73 73 494
mouton_valachie mouton 0 12 0 0 609 426 92 91 621
muntjac_chine muntjac 2000 49 0 0 0 0 0 0 2049
nandou nandou 1000 48 0 0 0 0 0 0 1048
oie oie 1000 73 0 0 0 0 0 0 1073
oryx_algazelle oryx 1196 74 0 0 0 0 0 0 1270
oryx_arabie oryx 852 79 0 0 0 0 0 0 931
ouistiti_pygmee ouistiti 1097 76 0 0 0 0 0 0 1173
ours_baribal ours 1132 69 0 0 0 0 0 0 1201
ours_lunettes ours 491 73 0 0 0 0 0 0 564
panda_roux panda 611 62 0 0 0 0 0 0 673
panthere_chine panthere 525 8 0 0 246 172 37 37 779
panthere_neige panthere 471 53 0 0 288 202 43 43 812
patas patas 1292 70 0 0 0 0 0 0 1362
pelican_blanc pelican 1000 64 0 0 0 0 0 0 1064
perruche_patagonie perruche 541 73 0 0 0 0 0 0 614
phacochere phacochere 1000 63 0 0 0 0 0 0 1063
pigeon_madagascar pigeon 976 11 0 0 0 0 0 0 987
raton_laveur raton_laveur 1000 70 0 0 0 0 0 0 1070
renard_polaire renard 1000 83 0 0 0 0 0 0 1083
rhinoceros_blanc rhinoceros 1000 81 0 0 0 0 0 0 1081
saimiris_perou saimiri 2034 68 0 0 251 176 38 37 2353
saki_face_blanche saki 1206 62 0 0 0 0 0 0 1268
serval serval 455 54 0 0 227 159 34 34 736
sitatunga sitatunga 319 48 0 0 286 200 43 43 653
springbok springbok 647 71 0 0 0 0 0 0 718
suricate suricate 478 73 0 0 0 0 0 0 551
tamarin_goeldi tamarin 249 47 0 0 160 112 24 24 456
tamarin_lion tamarin 407 5 0 0 276 193 42 41 688
tamarin_pinche tamarin 1288 92 0 0 0 0 0 0 1380
tamarin_roux tamarin 655 51 0 0 0 0 0 0 706
tapire_terrestre tapire 833 70 0 0 0 0 0 0 903
tigre_siberie tigre 1102 80 0 0 0 0 0 0 1182
titi_roux titi 159 0 0 0 257 180 39 38 416
urubu_tete_rouge urubu 1000 81 0 0 0 0 0 0 1081
vanneau_soldat vanneau 1000 36 0 0 0 0 0 0 1036
vautour_fauve vautour 1000 86 0 0 0 0 0 0 1086
vautour_ruppel vautour 1234 10 0 0 0 0 0 0 1244
vigogne vigogne 1000 94 0 0 0 0 0 0 1094
wallaby_bennet wallaby 1000 79 0 0 0 0 0 0 1079
watussi watussi 1000 101 0 0 0 0 0 0 1101
zebre_chapman zebre 696 76 0 0 0 0 0 0 772
TOTAL 117478 7521 0 0 9282 6498 1398 1386 134281

Sources

Source Folder Code Default size
iNaturalist inaturalist/ ina medium (500 px)
Wikimedia Commons wikimedia/ wkm medium (500 px)
Wikipedia wikipedia/ wkp medium (500 px)
Flickr flickr/ fli medium (500 px)
DuckDuckGo ddg/ ddg —

Filtering

  • iNaturalist – only Alive annotations (term 17 / value 18); quality grades research, needs_id, casual; wild and captive both kept (recorded in inat_savage). Image sizes restricted to small (240 px) or medium (500 px).
  • Flickr / Wikimedia / Wikipedia / DDG – negative-keyword title/URL filter (dead, carcass, skull, bones, taxidermy, trophy, hunt, feces, dung, scat, slaughter, roadkill, …).
  • DDG + Wikipedia – additional CLIP alive vs not-alive classification (openai/clip-vit-base-patch32); rejected URLs are persisted in <src>/rejected_urls/<worker>.csv so they are skipped on subsequent runs.

Repository layout 

Horama/wow_scraped/
├── inaturalist/
│   ├── 0-000.parquet               # worker 0, shard 0 (~500 MB)
│   ├── 0-001.parquet
│   ├── 3-000.parquet               # worker 3, shard 0
│   ├── …
│   ├── metadata.csv                # consolidated metadata
│   └── metadata/
│       ├── 0-000.csv               # per-shard metadata (raw)
│       └── …
├── wikimedia/  …
├── wikipedia/  …                   # also contains rejected_urls/
├── flickr/     …
└── ddg/        …                   # also contains rejected_urls/

Schema

Each parquet shard:

image:        struct<bytes: binary, path: string>   # HF Image feature
filename:     string    # src3_VV_W_NNNNNN.ext (e.g. ina_00_3_000042.jpg)
label:        string    # training label (e.g. loup_arctique)
specie:       string    # broader group (e.g. loup)
sub_specie:   string    # = label (full granularity)
src:          string    # inaturalist | wikimedia | wikipedia | flickr | ddg
licence:      string    # CC0, CC-BY, ..., unknown
author:       string
url:          string    # original source URL
resolution:   string    # "WxH"
split:        string    # train | valid | test (added by scraping.add_split_column)
inat_obs:     string    # iNaturalist observation id (other sources: "")
location:     string    # "lat,lon" when available, else ""
inat_quality: string    # research | needs_id | casual (iNat only)
inat_savage:  string    # "savage" | "captive" (iNat only)

Loading 

from datasets import load_dataset

# All shards from one source
ds = load_dataset("Horama/wow_scraped", data_dir="inaturalist", split="train")
print(ds[0]["image"])          # PIL.Image.Image
print(ds[0]["label"], ds[0]["licence"])

# A single shard
ds = load_dataset(
    "Horama/wow_scraped",
    data_files={"train": "inaturalist/0-000.parquet"},
    split="train",
)

Filtering by split

The split assignment (train / valid / test) lives in the per-source metadata.csv only — the parquets themselves don't carry it (so you keep the freedom to re-split without re-uploading the images). Join the metadata to filter:

import csv
from datasets import load_dataset
from huggingface_hub import hf_hub_download

split_lookup: dict[str, str] = {}
csv_path = hf_hub_download("Horama/wow_scraped", repo_type="dataset",
                            filename="inaturalist/metadata.csv")
with open(csv_path) as f:
    for row in csv.DictReader(f):
        split_lookup[row["filename"]] = row["split"]

ds = load_dataset("Horama/wow_scraped", data_dir="inaturalist", split="train")
train = ds.filter(lambda r: split_lookup.get(r["filename"]) == "train")
valid = ds.filter(lambda r: split_lookup.get(r["filename"]) == "valid")
test  = ds.filter(lambda r: split_lookup.get(r["filename"]) == "test")

iNaturalist observations are kept whole (no inat_obs is split between train/valid/test).

Licensing

Per-image license is preserved verbatim in the licence column so you can filter / partition the dataset for either research or production use.

Common values Allowed for
CC0, Public Domain Mark 1.0, No known copyright restrictions, US Government Work any (incl. commercial)
CC BY, CC BY-SA any with attribution
CC BY-NC, CC BY-NC-SA, CC BY-NC-ND, cc-by-nc research only
unknown (DDG) research only, attribution undetermined

Reproducing the dataset

The full pipeline (workers, sharding, alive filter, mailer) lives at Horama/WOW_dataset_creation under scraping/.

Citation

If you use this dataset, please credit Horama and the underlying contributors of each photo (see licence and author columns).
Downloads last month
81
Total file size:
16.6 GB

Collection including Horama/wildlife-scrap